January 1, 2016;
FoxH1 mediates a Grg4 and Smad2 dependent transcriptional switch in Nodal signaling during Xenopus mesoderm development.
In the vertebrate blastula
pathway is essential for formation of the primary germ layers and the organizer
autoregulatory feedback potentiates signaling activity, but mechanisms limiting embryonic Nodal
ligand transcription are poorly understood. Here we describe a transcriptional switch mechanism mediated by FoxH1
, the principle effector of Nodal
contains a conserved engrailed homology (EH1) motif that mediates direct binding of groucho-related gene 4
), a Groucho family corepressor. Nodal
-dependent gene expression is suppressed by FoxH1
, but enhanced by a FoxH1
EH1 mutant, indicating that the EH1 motif is necessary for repression. Grg4
-induced mesodermal gene expression and Nodal
autoregulation, suggesting that Grg4
pathway activity. Conversely, blocking Grg4
function in the ectoderm
results in ectopic expression of Nodal
target genes. FoxH1
occupy the Xnr1
enhancer, and Grg4
occupancy is dependent on the FoxH1
EH1 motif. Grg4
occupancy at the Xnr1
enhancer significantly decreases with Nodal
activation or Smad2
overexpression, while FoxH1
occupancy is unaffected. These results suggest that Nodal
physically displaces Grg4
, an essential feature of the transcriptional switch mechanism. In support of this model, when FoxH1
is unable to bind Smad2
occupancy is maintained at the Xnr1
enhancer, even in the presence of Nodal
signaling. Our findings reveal that FoxH1
mediates both activation and repression of Nodal
gene expression. We propose that this transcriptional switch is essential to delimit Nodal
pathway activity in vertebrate germ layer
Xla Wt + nodal1
[+] show captions
References [+] :
Fig. 1. FoxH1 contains a conserved EH1 motif that mediates a physical interaction with Grg4. (A) A diagram of the FoxH1 protein shows the location of the EH1 motif between the winged helix (WH) DNA binding domain and the Smad Interaction Domain (SID). This motif is highly conserved among multiple species (Xenopus, zebrafish, mouse and human). In the FoxH1A6 mutant, 6 of the 7 residues that comprise the EH1 motif have been mutated to alanine. (B) FoxH1 physically interacts with Grg4. mRNA encoding GST, GST-FoxH1 or GST-FoxH1A6 (1 ng) was injected along with myc-Grg4 mRNA (5 ng) into single cell embryos, which were collected at the gastrula stage for GST pulldown assays (Left). Top panel represents an anti-myc western blot, showing the physical interaction of myc-Grg4 with GST-FoxH1 (lane 3), which is reduced with mutation of the EH1 motif in FoxH1A6 (lane 4). Lower panel is an anti-GST western indicating recovery of GST proteins by pulldown. (Right) Control western blots showing approximately equal expression of myc-tagged and GST-tagged proteins in the starting lysates.
Fig. 2. FoxH1 represses Nodal target genes at blastula and gastrula stages. (A) FoxH1, but not FoxH1A6, represses expression of endogenous mesodermal and Nodal genes. Embryos were injected in the vegetal pole with mRNA encoding myc-FoxH1 or myc-FoxH1A6 (250 pg). cDNA was prepared from pooled whole embryos collected at blastula stage (for Xnr5 and Xnr6) and gastrula stage (for Xbra, Chd and Gsc). qPCR was performed for the mesodermal marker Brachyury, the organizer genes Chordin and Goosecoid, and the Nodal ligands Xnr1, Xnr5 and Xnr6. The data presented are the combined results of three independent biological replicates, and have been normalized to expression of a housekeeping gene, EF1α. Error bars represent standard error. * indicates p<0.05 by Student's t-test. (B) FoxH1, but not FoxH1A6, inhibits expression of Xnr5 and Xnr6 in the blastula vegetal pole. Embryos injected in the vegetal pole with mRNA encoding myc-FoxH1 or myc-FoxH1A6 (250 pg) were collected at the late blastula stage and subjected to in situ hybridization for either Xnr5 or Xnr6. (C) FoxH1, but not FoxH1A6, blocks expression of Xbra and Chd in the gastrula. Embryos were injected in the vegetal pole with mRNA encoding myc-FoxH1 or myc-FoxH1A6 (250 pg) and collected at the mid-gastrula stage. In situ hybridization was performed to analyze expression of Xbra and Chd. 100 embryos were analyzed for each condition, and the percentage of embryos displaying normal (dark gray bar) or reduced (light gray bar) expression was quantified in the bar graphs to the right. Expression of FoxH1 reduced the expression of Xnr5, Xnr6, Xbra and Chd.
Fig. 3. Grg4 inhibits Nodal-dependent mesoderm induction. (A) mRNA encoding Xnr1 (30 pg) or mRNA encoding myc-Grg4 (5 ng), or a combination of the two, were injected into the animal pole of one-cell stage embryos. Animal caps were prepared at blastula stage and analyzed for the expression of mesodermal and Nodal gene expression by RT-PCR at gastrula stage. Xnr1 induced expression of Xbra, Xwnt8 and the organizer gene Gsc, as well as Xnr1, Xnr2, Xnr4 and Der. Coexpression of myc-Grg4 blocked the upregulation of all these genes. Ef1α served as a control for RNA recovery and loading controls. Uninjected animal caps and PCR from a cDNA sample made without reverse transcriptase (Embryo-RT) showed no amplification. Whole embryo cDNA was used as a positive control. (B) Grg5 expression is sufficient to induce mesoderm in animal pole explants. mRNA encoding myc-Grg4 or myc-Grg5 (5 ng) was injected into the animal pole at the one-cell stage, and explants were isolated as described. Grg5 induced the mesodermal genes, as well as Xnr1 and Der. (C) Grg5 expression is sufficient to induce convergent extension. mRNA encoding myc-Grg4 or myc-Grg5 (5 ng) was injected into the animal pole of one-cell stage embryos. Explants were isolated at the blastula stage and allowed to develop until the neurula stage. Pictured is an unmanipulated embryo, which serves as a control for staging, and representative control, myc-Grg4, and myc-Grg5 injected explants. (D) Grg4 and Grg5 modulate the expression of a FoxH1-dependent reporter. The 3xARE luciferase reporter plasmid (100 pg) was injected along with a Renilla luciferase control plasmid (10 pg) at the one-cell stage, followed by single-blastomere injection at the two cell stage with mRNA encoding myc-FoxH1 or myc-FoxH1A6 (250 pg) alone, or in combination with myc-Grg4 or myc-Grg5 mRNA (5 ng). The 3xARE reporter alone served as a control for basal activity. Data shown represents four independent experiments and error bars represent standard error. * indicates p<0.05 as compared to FoxH1 alone. (E) HDAC inhibition induces Nodal and mesodermal gene expression in ectoderm. Animal explants prepared from blastula embryos were cultured for 2 h in media containing 2 mM sodium butyrate (Na Butyrate – gray bars) or 2 mM valproic acid (black bars). cDNA was prepared from treated and untreated caps and qPCR was performed to assay expression of the Xnr1, Xnr5, Xnr6, Xbra, and Chd. Gene expression is normalized to Ef1α and is shown as fold increase in expression over untreated caps. Error bars represent standard error in four independent experiments.
Fig. 4. Grg4 occupies the endogenous Xnr1 Enhancer through interaction with the FoxH1 EH1 motif. Occupancy at the Xnr1 enhancer was evaluated by chromatin immunoprecipitation (ChIP) and quantitative PCR (qPCR) of embryos injected with (A) 250 pg myc-FoxH1 or (B) 8 ng myc-Grg4. Immunoprecipitation using anti-myc antibody was also performed on uninjected embryos (Control). Each result shown represents three independent experiments. The white bars represent qPCR for genomic Xnr1 3′UTR as a control. (C) Genomic DNA fragments recovered by ChIP from embryos injected with myc-Grg4 alone (8 ng) or in combination with untagged FoxH1 or FoxH1A6 (250 pg) were evaluated by qPCR for the Xnr1 Intron 1 enhancer (gray bars). The white bars represent qPCR for genomic Ef1α as a control. The data shown represent three independent biological replicates. * indicates that p<0.05 as compared to myc-Grg4 alone.
Fig. 5. Grg4 occupies the Xnr1 enhancer in a Nodal-responsive manner. (A) ChIP for myc-FoxH1 alone or coexpressed with Xnr1 (myc-FoxH1+Xnr1) or (B) ChIP for myc-Grg4 alone or coexpressed with Xnr1 (myc-Grg4+Xnr1) was evaluated by qPCR for enrichment of the Xnr1 enhancer. (C) ChIP for endogenous Smad2/3 in uninjected embryos or embryos expressing 50 pg Xnr1 mRNA (+Xnr1) was evaluated by qPCR for the Xnr1 enhancer. Rabbit IGG ChIP serves as a negative control (IGG). (D) ChIP for myc-Grg4 alone, or coexpressed with Xnr1 (myc-Grg4+Xnr1) or Smad2GFP mRNA (myc-Grg4+Smad2GFP) was evaluated by qPCR for the Xnr1 enhancer. (E) ChIP for myc-Grg4 expressed alone or in combination with the following: FoxH1 (myc-Grg4+FoxH1), FoxH1 and Xnr1 (myc-Grg4+FoxH1+Xnr1), FoxH1∆SID (myc-Grg4+FoxH1∆SID) or FoxH1∆SID and Xnr1 (myc-Grg4+FoxH1∆SID+Xnr1) were evaluated by qPCR for the Xnr1 enhancer. The data shown represent three independent biological replicates. Immunoprecipitation using anti-myc antibody was performed on uninjected embryos (Control). White bars represent qPCR for genomic Xmlc2 or genomic Xnr1 3′UTR as additional negative controls. * indicates p<0.05.
supp Fig. 1: Expression levels of tagged FoxH1 proteins. (A) Immunocytochemistry for myc-tagged FoxH1 proteins demonstrates equivalent expression and nuclear localization of wild-type and mutant proteins in Xenopus embryos (250 pg mRNA injection for each construct). (B) Western blot showing near equal expression of myc-FoxH1 and myc-FoxH1A6 protein expression in embryos injected with 250 pg mRNA. The equivalent of 0.5 embryo is loaded in each lane. MAPK expression serves as loading control. (C) Western blots for myc-FoxH1 or myc-Grg4 after chromatin immunoprecipitation from embryos. A 20 µl sample of the eluate from ChIP of myc-FoxH1 or myc-Grg4 was combined with Laemmli buffer and boiled 30 min at 95 °C. The resulting sample was subjected to Western blot analysis using 9E10 antibody. The Western approximates 10% of the total immunoprecipitate.
Suppl Fig. 2: Grg4 specifically blocks Nodal-dependent mesoderm induction. (A) One-cell stage embryos were injected in the animal pole with mRNAs encoding Xnr1 (30 pg), eFGF (400 pg), myc-Grg4 (5 ng) or the combinations indicated. RT-PCR was performed on animal explants prepared at blastula stage and collected at gastrula stage. Expression of mesodermal genes Xbra and Xwnt8 is shown, with Ef1α as a positive control.
Suppl. Fig. 3: FoxH1, Grg4 and Grg5 modulate the activity of the Xnr1 intron 1 enhancer reporter. The Xnr1 enhancer luciferase reporter plasmid (100 pg) was injected along with a Renilla luciferase control plasmid (10 pg) at the one-cell stage, followed by single-blastomere injection at the two cell stage with mRNA encoding myc-FoxH1 or myc-FoxH1A6 (250 pg) alone, or in combination with myc-Grg4 or myc-Grg5 mRNA (5 ng). The Xnr1 enhancer reporter alone served as a control for basal activity. Data shown represent seven independent experiments and error bars represent standard error. * indicates p<0.05.
suppl Fig. 4: Smad2/3 occupies the endogenous Xnr1 enhancer in a FoxH1-dependent manner. (A) ChIP for endogenous Smad2/3 in uninjected embryos demonstrates Xnr1 enhancer enrichment (gray bars) without enrichment for the Xmlc2 enhancer (white bars). (B) ChIP for endogenous Smad2/3 in uninjected embryos or embryos injected with mRNA encoding FoxH1 or FoxH1∆SID mRNA (250 pg) was evaluated by qPCR for the Xnr1 enhancer. The white bars represent qPCR for the Xnr1 3′UTR control, which exhibits no interaction. The data presented represent three independent experiments. (C) ChIP for myc-p300 either alone or coexpressed with Xnr1 was evaluated by qPCR for the Xnr1 enhancer. The white bars represent qPCR for genomic Xmlc2 as a control. * indicates p<0.05 when compared to myc-p300 alone. The data represent four independent experiments.
Endodermal Nodal-related signals and mesoderm induction in Xenopus.